Climate change impacts on durum wheat in Sardinia

The high sensitivity of agriculture to climate conditions and the great uncertainty on the combined effects of increasing CO2 concentration and changes in temperature and rainfall patterns on agricultural crops, reveals the crucial importance of research in this field. In the past decades, several studies were conducted for assessing the individual effects of higher temperatures, elevated CO2 and changed rainfall patterns on crops. Recently, climate change studies have been focused on the interactions among changing factors for better evaluating the impacts of future climate and for implementing appropriate adaptation and mitigation strategies in agriculture. This study aimed (i) to assess impacts of potential climate change in ambient carbon dioxide (CO2) level on production and phenology for two of the most important varieties of durum wheat at four experimental sites in Sardinia, Italy, assuming different soils, climate conditions and management practices, and (ii) to provide guidelines for realistic adaptation strategies in a typical Mediterranean area. The CERES-Wheat model in combination with a stochastic Weather Generator (WG), recently developed for climate change studies (M&Rfi, available from: http://www.ufa.cas.cz/dub/wg/marfi/marfi.htm), was used to quantify the climate change impacts on wheat growth and production. Synthetic weather series representing possible future climates were generated by modifying the WG parameters according to a set of GCM based climate change scenarios. To account for the uncertainties in future climate development, we used a set of 9 climate change scenarios derived by the pattern scaling technique. In particular, three GCM-based (HadCM3, NCAR and ECHAM) standardised scenarios were multiplied by three values (low, middle, high) of changes in global mean temperature (ΔTG). The values of ΔTG were obtained by MAGICC climate model assuming various combinations of climate sensitivity (1.5, 3 and 4.5 °C) and four emission scenarios (SRES B1, B2, A1B, A2). For each experimental site, 27 climate change scenarios (9 scenarios for 3 future periods: 2025, 2050, 2075) were developed. The use of three CGMs allows accounting for differences in annual pattern of change in individual climatic characteristics. Prior to the climate change impacts analysis, we performed the calibration and validation of the CERES-Wheat model obtaining results that indicate good performance of the crop model at the study area. Then, the linkage of M&Rfi Weather Generator with CERES-wheat was validated by comparing statistical characteristics of crop model output simulated using observed weather data versus synthetic weather series (generated by weather generator) for the actual climate. Results showed no statistical differences, indicating that the use of CERES-Wheat coupled with the M&Rfi WG provides an efficient and reliable method for assessing the impacts on agricultural production. In the climate change impact assessment, we explored separately direct CO2 effect (known as a fertilization effect) and indirect CO2 effect (related to changed weather conditions) for the three future periods. In general, the results show that the indirect effect of CO2 concentration is negative. Considering the four experimental sites and the high GCM-based scenarios, crop yield will decrease by 2-6% for 2025, and by 10-18% for 2075. On the other hand, considering both direct and indirect effects of CO2 concentration, the wheat yield will increase by 5-7% for 2025 and by 16-21% for 2075. This means that the positive fertilisation effect of increased CO2 concentration could be sufficient to level out the negative impact of indirect effects. The results obtained for the analysis of climate change impacts on durum wheat in Sardinia indicate that the decrease of mean yield determined by increasing CO2 concentration in the atmosphere could be reduced opting to sow wheat earlier than the usual sowing date. An earlier planting date would produce an additional increase in wheat yield and could be a reliable and efficient adaptation strategy for wheat cultivation in this Mediterranean area.